MENU

Click here to

×

Are you sure ?

Yes, do it No, cancel

Noise Power Spectrum Analysis

E mckenzie*, D Gauntt, UAB Medical Center, Birmingham, AL

Presentations

(Sunday, 7/12/2020)   [Eastern Time (GMT-4)]

Room: AAPM ePoster Library


Introduction:

Noise power spectrum (NPS) is a measurement used to characterize the texture of an imaging systems noise output in terms of spatial frequency. The goal of this experiment is to create a coded comparison tool useful for quantizing noise texture between different CT reconstruction kernels and providing the peak spatial frequency for each of them. This would be a useful tool for anyone with a desire to know the similarities between reconstruction kernels of different CT manufacturers.

Methods and Materials:

This experiment compares the NPS of various filter kernels from a scanned slice of the Gammex 464 ACR phantom on the GE Discovery 750HD, the Philips Brilliance 64 and the Siemens Somatom Force to provide guidance in matching image reconstruction filters. This work extends the work of Samei et al in their 2012 publication on comparison of noise texture between the GE Discovery CT 750HD and Siemens Somatom Definition Flash. However, in contrast to previously published work, this experiment subtracts background from the same location of a different image as opposed to subtracting from a quadratic function representing background made from various images, and extends the analysis to additional scanner models.

Results:

This experiment presents tables listing the best match for filter kernels between various pairs of scanner models, as well as peak frequency and predicted noise at constant CTDI for selected filter kernels.

Conclusion:

This analysis provides a useful tool for harmonizing CT protocols across scanner models of multiple vendors.

Keywords

Noise Power Spectrum, CT

Taxonomy

IM- CT: Quantitative imaging/analysis

Contact Email